USB and COM port share one command buffer
@Yukti007 ya está trabajando en esto.
Desde el 13/10/2026.
Evaluación
Este issue todavía no se ha evaluado.
Descripción
Describe the bug
The Krake reads text commands from two ports: the USB serial port and the RS-232 COM port. Both ports put their characters into the same command buffer, with the same character counter. If one port has sent part of a command but not the end-of-line yet, characters from the other port get added to the end of it. The Krake then runs one wrong command made from pieces of both.
Short Summary of the expected behavior:
Each port's commands are handled separately. A command from one port is never mixed with characters from the other port.
Short Summary of the buggy behavior:
A half-finished line from one port gets finished by characters from the other port. This can give the wrong alarm level, the wrong message text, or a lost alarm.
To Reproduce
Found by reading the code. Not yet tested on a real Krake board. Suggested steps to confirm:
- Connect a Krake to a PC by USB and open a serial monitor at 115200 baud, with line ending set to "No line ending".
- Connect a second serial terminal to the COM port (DB9, RS-232).
- On USB, type
a2Pumpand send it (no newline is sent). - On the COM port, send
a5Firewith a newline. - See error: the Krake shows a level 2 alarm with the text "Pumpa5Fire" instead of a level 5 alarm with the text "Fire".
Expected behavior
The COM port command a5Fire should give a level 5 alarm with the text "Fire". The unfinished text on USB should stay with the USB port and not affect the COM port.
Screenshots
None. Not yet tested on hardware.
Hardware description
Model: Krake (GPAD_API firmware), checked against main at commit b6ea4b3 (2026-09-20)
Serial Number: N/A (found by reading the code)
Other: Affects any Krake with both the USB serial and the COM port connected.
Test Equipment
Model: N/A (not yet tested)
Serial Number: N/A
Additional context
Where in the code:
- One shared buffer:
Firmware/GPAD_API/GPAD_API/gpad_serial.cpplines 34–35
https://github.com/PubInv/krake/blob/b6ea4b3eaf0e93a580cad7cdc9f8c1541860ff06/Firmware/GPAD_API/GPAD_API/gpad_serial.cpp#L34-L35 - One shared character counter (
static size_t writeIndex):gpad_serial.cppline 54
https://github.com/PubInv/krake/blob/b6ea4b3eaf0e93a580cad7cdc9f8c1541860ff06/Firmware/GPAD_API/GPAD_API/gpad_serial.cpp#L54 - Both ports use the same function, one after the other, in
loop():GPAD_API.inolines 2591 and 2595
https://github.com/PubInv/krake/blob/b6ea4b3eaf0e93a580cad7cdc9f8c1541860ff06/Firmware/GPAD_API/GPAD_API/GPAD_API.ino#L2591-L2595
Why it happens: processSerial() keeps a half-finished line in the buffer between calls, but there is only one buffer and one counter for both ports. Nothing records which port the half-finished line came from. It is not two tasks running at once: everything runs in loop(), one call after the other.
Side effect: if one port sends a line that is too long, the buffer is cleared. That also throws away the other port's half-finished command and cancels pending alarm audio (gpad_serial.cpp lines 107–113).
How likely: low to medium. Most setups use one port at a time, but it is easy to hit while debugging with a serial monitor open.
Possible fix: give each port its own buffer and its own character counter, passed into processSerial(). The comment at gpad_serial.cpp lines 45–46 already suggests making the buffer a parameter.
Related: #283 (shared global variables).
- Lenguaje dominante
- C++
- Estrellas
- 9
- Forks
- 6
- Métricas de merge de PR
- Sin PR fusionados en 30 d
Preparar el entorno
- Sin Dockerfile ni archivo de Docker Compose
- Sin plantilla de pull request
- Leer la guía de contribución
Primeros pasos
- Lee el issue completo y luego la guía de contribución del proyecto.
- Comenta en el issue que vas a ocuparte — evita que dos personas hagan lo mismo.
- Haz un fork del repositorio y trabaja en una rama.
- Abre un pull request que haga referencia al número del issue.
Más de PubInv/krake
-
web page publish and main loop share the MQTT send queuePosiblemente ocupada @Yukti007 la tomó hace 13 días. AbiertoHigh Priority regression test type: bug
-
Firmware, Krake Loosing Broker Settings back to a DefaultPosiblemente ocupada @nk25719 la tomó hace 25 días. Abiertotype: bug
-
Hardware, LB0007 Not Playing Alarm FileQuizá libre de nuevo @nk25719 la tomó hace 32 días y no hay ningún pull request abierto. Abierto
-
Mini MP3 File Names and File Order For Reliable Play Back, Both GPAD_API and Factory TestQuizá libre de nuevo @Yukti007 la tomó hace 34 días y no hay ningún pull request abierto. Abiertodocumentation
-
Dificultad 5/5 Más de una semana Aptitud para principiantes 15/100
Todos los issues de PubInv/krake
Issues similares
-
[Bug] [Go] Ack returns nil when the server answers with a non-OK statusPosiblemente ocupada @leizhiyuan la tomó hoy. Abierto
Dificultad 2/5 1-3 horas Aptitud para principiantes 84/100
apache/rocketmq-clients#1465 ·
Los mantenedores suelen responder en 2 días
-
Dificultad 2/5 1-3 horas Aptitud para principiantes 72/100
KhronosGroup/Vulkan-Tutorial#534 ·
-
Dificultad 2/5 1-3 horas Aptitud para principiantes 65/100
Los mantenedores suelen responder en 2 días
-
bug IO
Dificultad 2/5 1-3 horas Aptitud para principiantes 65/100
MrNeRF/LichtFeld-Studio#3824 ·
Los mantenedores suelen responder en 1 día
-
pnnx: manual tanh GELU fusion loses the approximation modePosiblemente ocupada @ktz03 la tomó hoy. Abierto
Dificultad 2/5 1-3 horas Aptitud para principiantes 78/100
Tencent/ncnn#7175 · 1 reacción ·
Los mantenedores suelen responder en 1 día